A distinct subset of known oncogenes, including wild-type alleles of c- MYC or GLI, and activated alleles of RAS or beta-catenin, are capable of transforming cultured diploid epithelial cells in vitro (RK3E cells). By expression cloning we normally confined to the non-proliferating, superficial cell layers. Unlike for colorectal carcinoma, expression of GKLF expression epithelia as normally confined to the non-proliferating, superficial cell layers. Unlike for colorectal carcinoma, expression of GKLF mRNA and proteins are greatly increased early during progression of most cases of ductal carcinoma of the breast and oral squamous cell carcinoma. These tumor-types are not frequently associated with activation of other known oncogenes such as beta-catenin, RAS, or GLI. Deregulation of GKLF mRNA is therefore attractive as a candidate mechanism of malignant transformation and tumor progression in breast cancer. Expression of integrin mRNA and protein is reduced in dysplastic epithelium and in tumors, and reduced expression of integrins in breast tumors as a predictor of metastasis to axillary lymph nodes. Integrins are important for polarization of epithelial cells, and restoration of integrin expression in breast cancer cell lines restores multiple characteristics of normal breast epithelia. GKLF inhibited expression of the integrins beta1 and alpha3, encoding a basement membrane receptor, but activated the vitronectin receptor alphavbeta3, previously implicated in the metastatic phenotype of breast cancer cells. Expression of a GKLF transgene in the basal cell layer of mouse skin induces loss of the apical-basal polarization of these cells, increased cellularity, and other features of dysplasia. We will use a novel monoclonal antibody to characterize expression of GKLF in pre-invasive lesions and in invasive ductal carcinoma of the breast. GKLF expression will be correlated with expression of other prognostic markers, including integrins, and with clinical and pathological parameters including survival (Aim 1). We will test whether increased expression of GKLF in breast epithelium of transgenic mice induces atypia or predisposes mice to tumor progression (Aim 2). Finally, we will use RK3E epithelial cells and human mammary epithelial cells to further examine the mechanism and significance of integrin regulation by GKLF (Aim 3).